Diagnostic testing is a universal feature of modern life, spanning medicine, engineering, finance, software, and government. Yet across these domains, a large fraction of tests fail to influence any subsequent decision, generating vast economic and human inefficiencies. This paper introduces a unified decision‑theoretic principle: a test has value only if at least one possible outcome leads to an action the decision‑maker is willing to take. When this condition is not met, testing produces information without consequence—data that cannot alter behavior. By analyzing major global sectors, we show that non‑actionable testing generates an estimated 1.5 to 2 trillion USD in direct annual waste, with an additional 500 billion USD in lost human time. Together, these inefficiencies represent roughly two percent of global GDP. The paper formalizes the minimal condition for rational testing, explains why real‑world systems routinely violate it, and provides a practical rule that individuals and institutions can apply to eliminate low‑value testing and restore the link between information and meaningful action.
Florin Horicianu (Sun,) studied this question.